html-report
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection by generating HTML documents based on processed data. If the input data is sourced from untrusted external locations, malicious content could be embedded in the final HTML report. * Ingestion points: The data variable in the Python snippet in SKILL.md. * Boundary markers: Uses json.dumps for structural delimitation but lacks HTML body sanitization in SKILL.md. * Capability inventory: Python's open().write() to the workspace in SKILL.md. * Sanitization: Relies on json.dumps(data, ensure_ascii=False) in SKILL.md.
- [EXTERNAL_DOWNLOADS]: The skill allowlists specific CDNs (cdnjs.cloudflare.com, cdn.jsdelivr.net, unpkg.com, esm.sh, fonts.googleapis.com, fonts.gstatic.com) for assets in SKILL.md.
- [DYNAMIC_EXECUTION]: The skill generates self-contained HTML reports containing JavaScript and CSS from templates, which are then written to the task directory in SKILL.md.
Audit Metadata